In those cases in which the maximum rate corresponding to the optimum concentration of phosphate is never attained, some secondary cause may be supposed to intervene, such as a permanent change in a portion of the fermenting complex, accumulation of the products of the reaction, etc. It is also possible as suggested by Buchner for the analogous case of arsenite (p. 78) that the addition of increasing amounts of phosphate causes a progressive but reversible change in the mode of dispersion [p073] of the colloidal enzyme, and that this has the secondary effect of altering the rate of fermentation. No decisive evidence is as yet available upon the subject. The results obtained by Euler and Johansson [1913] to which reference has already been made indicate that in presence of a moderate excess of phosphate esterification is more rapid than production of carbon dioxide. No explanation of this phenomenon has yet been given, but it might obviously be due either to the production of some phosphorus compound which subsequently takes part in the production both of hexosediphosphate and of carbon dioxide, or, less probably, to the entire independence of the two changes--esterification of phosphate and production of carbon dioxide--which might then be differently affected by the presence of excess of phosphate and therefore take place at different rates. II. REACTION OF FRUCTOSE WITH PHOSPHATES IN PRESENCE OF YEAST-JUICE. Although, as has been pointed out (p. 42), glucose, mannose, and fructose all react with phosphate in a similar manner in presence of yeast-juice, there are nevertheless certain quantitative differences between the behaviour of glucose and mannose on the one hand, and fructose on the other, which appear to be of considerable importance. Fructose differs from the other two fermentable hexoses in two particulars: (1) the optimum concentration of phosphate is much greater; (2) the maximum rate of fermentation attainable is much higher [Harden and Young, 1908, 2; 1909]. These points are clearly illustrated by the following results, which all refer to 10 c.c. of yeast-juice, and show that the optimum concentration of phosphate for the fermentation of fructose is from 1·5 to 10 times that of glucose, and that the maximum rate of fermentation for fructose in presence of phosphate is 2 to 6 times that of glucose.